Cyclo Smart Bike: Bicycle that Prevent Collision and Hazardous Location Detection Through the Integrating of Multiple Sensor

Thean, Lee Hong (2025) Cyclo Smart Bike: Bicycle that Prevent Collision and Hazardous Location Detection Through the Integrating of Multiple Sensor. Final Year Project (Bachelor), Tunku Abdul Rahman University of Management and Technology.

Text Thean Lee Hong - FULL TEXT.pdf Restricted to Registered users only Download (1MB)

Abstract

Cyclist safety remains a critical concern due to the increasing risk of collisions with rear-approaching vehicles, often made worse by poor visibility and blind spots. This paper presents a novel system to enhance cyclist safety by designing a framework of collision detection system on a bicycle to detect and track vehicles approaching from behind. The system utilizes Density-Based Spatial Clustering of Applications with Noise (DBSCAN) to cluster LiDAR data points for identifying potential vehicles and employs an Extended Kalman Filter (EKF) to estimate their positions and velocities, with a distinctive feature of auto-tuning the EKF noise parameters based on innovation error through an Adaptive EKF (AEKF), thereby improving estimation accuracy under varying conditions. Extensive simulations demonstrate the system’s ability to accurately detect and track multiple rear vehicles, achieving high estimation errors accuracies and enabling timely cyclist alerts. This work advances cyclist safety technology by integrating with adaptive algorithms, offering a scalable solution to reduce bicycle-related accidents.